We propose three related area of research, each focused on a major aspect of Ras signaling. The first attempts to solve an important question relating to the NF1 gene, now recognized as a major tumor suppressor in human cancer as well as genetic basis for neurofibromatosis, a disease that affects 1 in 3500 people worldwide. Using biochemical approaches as well as cell-based screens, we will try and identify proteins or other cellular factors that regulate neurofibromin's ability to regulate Ras. Recently, we identified the first bone fide neurofibromin binding proteins other than Ras, the Spred proteins that are essential for neurofibromin's activity. We will build on this discovery to determine how the neurofibromin/Spred complex is assembled and regulated, and, in addition, we will perform unbiased screens for factors that regulate Ras through neurofibromin, or regulate Ras-independent functions of this highly conserved protein. Second, we will analyze individual mutants of K-Ras for their requirement for upstream signaling, and for the specific activation of downstream pathways. We will do this, as in the previous section, by combining analysis of biochemical properties of the proteins themselves, and analysis of downstream pathways. We will focus on known pathways that are clearly activated differently by different oncogenic forms of Ras, such as the Raf MAPK pathway, and unbiased analysis of signaling differences that are revealed by analysis of individual proteins in a clean, isogenic background. Finally, we have discovered that K-Ras promotes a stem-cell like phenotype in cancer cells that enables them to form tumors very efficiently from small numbers of cells, to metastasize and to become drug resistant. These phenotypes are caused by K-Ras binding to calmodulin and suppressing non-canonical wnt signaling, while increasing transcription from beta-catenin and other factors involved in stem-cell signaling. One of these factors that is produced by K-Ras cancers and plays a major role in the stem-cell phenotype is Leukemia Inhibitory Factor (LIF). We have found that neutralizing LIF prevents tumor initiation in vivo and sensitizes pancreatic tumor cells to gemcitabine, so that established tumors can be completely eradicated. We propose to undertake pre-clinical analysis of humanized anti-LIF monoclonal antibodies, using several models of pancreatic cancer, with the intention of moving these antibodies into clinical testing based on the results of these experiments. Binding of K-Ras to calmodulin is prevented by phosphorylation on serine-181, by PKC. The non-tumor promoting natural product, prostratin, activates PKC in vivo and disrupts K-Ras-calmodulin interaction. Oral administration of prostratin prevents development of pancreatic cancers in mouse models, with no obvious side effects. Several analogs of prostratin have been synthesized: we will test all of these, and make more. We will identify which form of PKC is involved in K-Ras phosphorylation and perform thorough preclinical testing for safety and efficacy of the most promising analogs, so that, if successful, we will advance one of these towards clinical testing.